Prediction of the three-dimensional structure of protein associations with components that undergo conformational deformations and partial re-structuring upon binding is a great challenge. It represents a major bottleneck in the modern structural understanding of biological function and disease. Ensemble docking emerged as a practical approach for incorporating conformational variability of a part of the system. Innovative ways to generate the ensembles using systematic omission scans and/or local relevant normal modes were developed recently by our group and tested in small ligand docking. The methods showed encouraging results in previously unsolvable cases, and are directly transferable to protein docking. In the present proposal, these methods will be extended and applied to protein docking including the most difficult docking of fully unstructured isolated or terminal peptides. Firstly, the conformational ensemble approach will be dramatically accelerated by introducing a new 4D docking procedure in which atomic models with fully flexible parts will be docked in a single run into concurrently present multiple conformation fields. Secondly, a faster and more rigorous all-atom solution refinement protocol will be applied. This protocol operates on softly restrained and fully flexible interface patches. All methods will be tested on a comprehensive induced fit benchmark for protein and peptide interactions that will be made publically available and regularly updated. Finally, the proposed docking protocol will incorporate electron microscopy (EM) data and other experimental restraints. A new damped dynamics flexible fitting method designed for EM fitting will be further developed. The new protein and peptide docking methods and multi domain EM fitting methods will be applied to solving biological problems with collaborating experimental laboratories. Structure prediction of protein and peptide complexes will lead to the discovery and characterization of new sites that can be targeted with small molecule therapeutics.